Task-Driven Subspace Decomposition for Knowledge Sharing and Isolation in LoRA-based Continual Learning
arXiv:2603.00191v1 Announce Type: new Abstract: Continual Learning (CL) requires models to sequentially adapt to new tasks without forgetting old knowledge. Recently, Low-Rank Adaptation (LoRA), a …
Lingfeng He, De Cheng, Huaijie Wang, Xi Yang, Nannan Wang, Xinbo Gao
19 views